On Wed, 30 Jan 2019, Sebastian Sewior wrote:

> On 2019-01-30 13:59:55 [+0100], Heiko Carstens wrote:
> > Last lines of trace below (full log attached):
> 
>             <...>-56956 [005] ....   658.931364: handle_futex_death: uaddr: 3ff9e880c58 pi: 1
> …
>             <...>-56956 [005] ....   658.931369: handle_futex_death: uaddr: 3ff9e8808c0 success
>             <...>-56956 [005] ....   658.931370: sched_process_exit: comm=ld64.so.1 pid=56956 prio=120
> 
> not including 3ff9e880140
> 
>             <...>-56956 [005] ....   658.931370: sched_process_exit: comm=ld64.so.1 pid=56956 prio=120
> 
>             <...>-56496 [001] ....   658.932404: sys_futex(uaddr: 3ff9e880140, op: 6, val: 1, utime: 0, uaddr2: 5, val3: 0)
>             <...>-56496 [001] ....   658.932405: attach_to_pi_owner: Missing pid 56956
>             <...>-56496 [001] ....   658.932406: handle_exit_race: uval2 vs uval 8000de7c vs 8000de7c (-1)
>             <...>-56496 [001] ....   658.932501: sys_futex -> 0xfffffffffffffffd
> …
>             <...>-56496 [001] ....   659.020750: handle_futex_death: uaddr: 3ff9e880140 pi: 1
>             <...>-56496 [001] ....   659.020750: handle_futex_death: uaddr: 3ff9e880140 uval: 8000de7c
>             <...>-56496 [001] ....   659.020750: handle_futex_death: uaddr: 3ff9e880140 success
> 
> and here we have it.

Bah. I was looking at the wrong trace. /me goes to give a talk and stares
at that mess later.